- Email: [email protected]
Eosinophil-derived interferon-lambda contributes to initiation of allergen-related inflammation in the intestine
Cytokine 58 (2012) 186–192
Contents lists available at SciVerse ScienceDirect
Cytokine journal homepage: www.elsevier.com/locate/issn/10434666
Eosinophil-derived interferon-lambda contributes to initiation of allergen-related inflammation in the intestine Shao-Heng He a,⇑,1, Chun-Hua Song b,1, Zhigang Liu c, Huiyun Zhang a, Wenjing Ma a, Lin-Fu Zhou a, Tahrin Mahmood b, Ping-Chang Yang b,⇑ a b c
Clinical Experimental Center, The First Affiliated Hospital, Nanjing Medical University, Nanjing 210029, China Department of Pathology & Molecular Medicine, McMaster University, Hamilton, ON, Canada Allergy and Immunology Institute, School of Medicine, Shenzhen University, Shenzhen, China
a r t i c l e
i n f o
Article history: Received 9 November 2011 Received in revised form 22 December 2011 Accepted 9 January 2012 Available online 4 February 2012 Keywords: T helper 2 response Eosinophil Interferon lambda Epithelial barrier Food allergy
a b s t r a c t Background and aims: Epithelial barrier dysfunction plays a critical role in the initiation of a number of immune diseases; the causative factors are not fully understood. The present study aimed to elucidate the mechanism by which the eosinophil-derived interferon (IFN)-lambda induced the gut epithelial barrier dysfunction. Methods: The duodenal biopsies were obtained from patients with or without food allergies. The eosinophils and IFNk expression were observed by immune staining. Intestinal epithelial cell line, T84 cells, and a mouse model were employed to observe the effect of IFNk on the epithelial barrier function and the initiation of skewed T helper (Th)2 polarization in the mouse intestine. Results: IFNk expression was observed in over 80% human eosinophils of the subjects with or without food allergies. Exposure to microbial products, lipopolysaccharide or peptidoglycan, could induce eosinophils to release IFNk. Exposure to IFNk could induce intestinal epithelial barrier dysfunction via inducing the epithelial cell apoptosis. Concurrent exposure to microbial products and food antigens could induce aberrant antigen specific Th2 polarization and Th2 pattern inflammation in the intestine. Conclusions: Eosinophils express IFNk that can induce intestinal epithelial barrier dysfunction and promotes the initiation of the aberrant Th2 polarization in the intestine. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction The antigen absorption in the intestinal mucosa is an essential step to initiate immune responses in the intestine. Under physiological conditions, the absorption of antigens is restricted by the integrated epithelial barrier on the intestinal mucosal surface. However, the integration of the intestinal epithelial barrier can be compromised by a number of factors, such as under psychological stress [1,2], or severe microbial infection [3,4], etc. These conditions may cause the tight junction open, increase in the intracellular permeability, or epithelial cell apoptosis resulting in the physical defect in the epithelial barrier; yet, the mechanisms underlying these conditions are not fully understood. Food allergy indicates a condition that the immune system in the intestine aberrantly responds to normal food components.
⇑ Corresponding authors. Address: Room T3303, 50 Charlton Ave. East, Hamilton, ON, Canada L8N 4A6. Tel.: +1 905 522 1155x32934; fax: +1 905 540 6593 (P.-C. Yang), tel./fax: +86 25 83718836 (S.-H. He). E-mail addresses: [email protected] (S.-H. He), [email protected] (P.-C. Yang). 1 These authors contributed equally to this work. 1043-4666/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.cyto.2012.01.003
The pathological feature of food allergy is a skewed T helper (Th)2 polarization, elevated levels of antigen specific IgE, mastocytosis and eosinophilia in the intestinal mucosa [5]. The initiation of the skewed Th2 polarization is not fully understood yet; multiple factors may be involved in the induction of food allergy. Interferon (IFN)k is a newly described cytokine, including IFNk1 [interleukin (IL)-29], IFNk2 (IL-28A) and IFNk3 (IL-28B) three subtypes; the IFNk1 is only expressed in human. IFNk is actively involved in the immune regulation, such as inducing target cell apoptosis [6,7], suppressing IL-13 expression [8], facilitating regulatory T cell development [6,9] and fighting against viral infections. It has been found that IFNk can be produced by several cells types, including epithelial cells [10], dendritic cells [6] and macrophages [11]. It is still not fully understood how the production of IFN initiated by those cells. Eosinophils (Eos) are an inflammatory cell type. The eosinophilia is found in a number of diseases, such as allergic disorders, parasite infections and some special inflammations like eosinophilic esophagitis [12]. Apart from producing a number of chemical mediators, such as the major basic protein (MBP), eosinophil cationic protein, eosinophil peroxidase, etc., we also noted that Eos also produced IFNk in our previous study ([6] data not shown);
S.-H. He et al. / Cytokine 58 (2012) 186–192
187
Fig. 1. Eos express IFNk in the intestine. Duodenal biopsies were processed for immunohistochemistry. The representative confocal images show that the Eo marker, MBP, was stained in green; the IFNk was stained in blue. The light blue color (pointed by arrows) was merged by the colors of green and blue. The nuclei were stained in red. (A) Samples were obtained from 15 patients with food allergy. (B) Samples were obtained from 15 non-food allergy patients. (C) An isotype control. (D) The bars indicate the double positive cell counts (mean ± SD). ⁄p < 0.01, compared with food allergy (FA) group. nFA, non-FA. Original magnification: 200. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
the phenomenon has not been reported yet as we checked the literature throughout the PubMed database. Since Eos are one of the effector cell types in allergic diseases, Eos express Toll like receptor (TLR)2 and TLR4 [13], we postulated that the Eo-derived IFNk might be involved in the initiation of the skewed Th2 polarization in the intestine. The present data show that upon exposure to the ligands of TLR2 or TLR4, Eos release IFNk; the latter affects the integrity of intestinal epithelial barrier function that contributes to the initiation of aberrant Th2 polarization and Th2 pattern inflammation in the intestine. 2. Materials and methods 2.1. Patients Thirty patients with peptic ulcer with or without food allergies were included in the present study. The diagnosis of food allergy is described in Supplementary materials. The patients did not use any anti-allergy drugs at least 1 month before the study. The demographic data of patients are presented in Supplementary Table 1.
2.4. Immunohistochemistry and TUNEL staining Duodenal biopsies were frozen in liquid nitrogen and stored at 80 °C until use. Cryosections were prepared and fixed with cold acetone for 20 min. Sections were blocked by 1% bovine serum albumin (BSA) for 30 min and incubated with goat anti-human IFNk antibody (1:200) and mouse anti-MBP antibody (1:300), or isotype IgG (using as negative control) overnight at 4 °C. After washing, sections were incubated with Cy5-labeled anti-goat and FITC-labeled anti-mouse second antibodies (1:300) for 1 h at room temperature. After washing, sections were mounted with cover slips and observed under a confocal microscope. The positive cells were counted in 20 fields (200) per samples. For T84 cell staining, the cell suspension was smeared onto glass slides, dried overnight. The rest procedures were the same as above. To avoid the observer bias, the sections were coded; the observer was not aware of the codes. After the MBP labeling, some sections were incubated for 1 h in freshly prepared TUNEL mixture containing TUNEL plus fluorescein tag in a dark humidified chamber at 37 °C. After washing, sections were mounted with coverslips and viewed under a confocal microscope.
2.2. Duodenal biopsy 2.5. Detecting intestinal epithelial barrier function in vivo Duodenal biopsies were taken under endoscopy for a purpose of ruling out malignant disorders in the duodenum. One piece of biopsies was taken from the edge of the ulcer; another piece was taken from non-ulcer mucosa. The procedures using human tissue in the study were approved by the Research Ethic Committee at Nanjing Medical University. Informed written consent was obtained from each subject.
BALB/c mice with eosinophilia were gavage-fed with ovalbumin (OVA) 5 mg/mouse in 0.3 ml saline. Mice were also i.p. injected with lipopolysaccharide (LPS; see Fig. 3C for dosage) or peptidoglycan (PGN; see Fig. 3D for dosage). Mice were sacrificed 6 h later. The jejunum segments were excised to extract the total proteins. The proteins were analyzed by Western blotting for the contents of the OVA.
2.3. Enzyme-linked immunoassay (ELISA) 2.6. Eosinophil depletion Levels of serum specific IgE, intestinal IL-4 and IFNk in tissue extracts and culture supernatant were measured by ELISA with commercial reagent kits following the manufacturer’s instruction.
Rats were i.p. injection with anti-CCR3 antibody (0.5 mg/rat) on day 3, 8 and 11, respectively. As shown by peripheral blood smear
188
S.-H. He et al. / Cytokine 58 (2012) 186–192
85% ± 11.8% of the total MBP+ cells in food allergy group and 82.1% ± 14.8% of the total MBP+ cells in non-food allergy group (p > 0.05; Fig. 1). In addition, similar results were also observed in human tissues of the skin, tonsil, colon, lung, kidney and stomach (data not shown). To confirm the data that Eos express IFNk, we infected mice with parasites to induce the eosinophilia in the intestine. The nucleated cells were isolated from the jejunum and examined by flow cytometry. The results showed that the rate of MBP+ IFNk+ cells was 92.5% ± 15.6% in the total MBP+ cells from six mice with parasite infection and 91.1% ± 16.6% in naïve mice. The data indicate that the number of Eo is significantly more in the duodenal mucosa of patients with food allergy than those of non-food allergy. Over 80% Eos in human and over 90% Eos in mice express IFNk. 3.2. Eos release IFNk upon activation by TLRs Previous reports indicate that Eos express TLR2 and TLR4 [13]. We next examined the expression of TLRs on the Eo cell line, EoL1 cells. As shown by flow cytometry, over 90% EoL1 cells express both TLR2 and TLR4 (Fig. 2A and B). The EoL1 cells also expressed IFNk (Fig. 2C and D). We then added the ligands of TLR2 (peptidoglycan; PGN; 100 ng/ml) [14] and TLR4 (lipopolysaccharide; LPS; 100 ng/ml) to activate the EoL1 cells in culture for 3 days. As assessed by ELISA, both TLR2 and TLR4 ligands could respectively induce the release of IFNk from EoL1 cells (Fig. 2E). 3.3. Intestinal epithelial cells express the IFNk receptors (IFNkR)
Fig. 2. EoL1 cells release IFNk upon TLR ligand stimulation. (A and B) The dot plots show TLR2 and TLR4 expressing EoL1 cells. (B) An isotype control. (C and D) The histograms indicate the expression of IFNk by the EoL1 cells in (A) and (B) (pointed by arrows). (E) The bars indicate the levels of IFNk in culture supernatants of EoL1 cells stimulated by LPS of PGN. The data were expressed as mean ± SD from three experiments. ⁄p < 0.01, compared with the dose ‘‘0’’ group.
and jejunal sections, eosinophils were observed in naïve control rats; no eosinophils were observed in rats treated with anti-CCR3 antibody. 2.7. Statistics All values were expressed as the means ± SD of at least three independent experiments. The values were analyzed using the two-tailed unpaired Student’s t-test when the data consisted of two groups or by ANOVA when three or more groups were compared. p < 0.05 was accepted as statistically significant. Other experimental procedures were listed in Supplementary materials. 3. Results 3.1. Eos express IFN in the intestine In the first attempt, we observed the expression of IFNk by Eos in the duodenal biopsies. As shown by immunohistochemistry, Eos were observed in the duodenal biopsy sections of the patients with or without food allergies. The number of Eos was significantly more in the patients with food allergies than that in non-food allergy (p < 0.01). Three types of positive cells, MBP+, IFNk+ and MBP+ IFNk+, were observed. The rate of MBP+ IFNk+ cells was
Eos are dispersed immediately under the subepithelial region as well as in the epithelial layer. Such an anatomical relation between Eos and the epithelium implies that the Eo-derived IFNk may affect the epithelial cells. By immunohistochemistry, the expression of IFNkR was observed on the epithelial cells of human duodenal biopsies as well as on the human intestinal cell line, T84 cells (Fig. 3). 3.4. Eo-derived IFNk induces gut epithelial barrier dysfunction The data in Figs. 2 and 3 imply that Eo-derived IFNk may affect the intestinal epithelial barrier function. To test the hypothesis, we cultured the human intestinal epithelial cell line, T84 cell monolayers, in transwells to confluence. EoL1 cells were added to the basal chambers of transwells in the presence of LPS or PGN. The results showed that upon the exposure to LPS or PGN, the transepithelial electric resistance (TER) was markedly dropped and the permeability to HRP was significantly increased, which could be blocked by addition of the anti-IFNk antibody (Fig. 4A and B). The data imply that the Eo-derived IFNk can compromise the intestinal epithelial barrier function. To further confirm the results in Fig. 4A, we induced the eosinophilia in mouse intestine by infecting mice with parasites. The mice with eosinophilia were injected with LPS or PGN via i.p. to induce the Eos to release IFNk; the mice were concurrently gavage-fed with OVA to determine the epithelial barrier permeability. As shown by Western blotting (Fig. 4C and D), the contents of OVA in the protein extracts of intestinal tissue were significantly increased in a TLR ligand dose-dependent manner, which could be blocked by pretreatment with the anti-IFNk antibody, but not by pretreatment with an isotype antibody (data not shown). The results indicate that Eo-derived IFNk can cause intestinal epithelial barrier dysfunction. To find the mechanism by which Eo-derived IFNk induces intestinal epithelial barrier dysfunction, we examined the T84 cells in Fig. 3 and the intestinal epithelial cells of mice in Fig. 4. The results showed that abundant apoptotic cells were detected among the
S.-H. He et al. / Cytokine 58 (2012) 186–192
189
Fig. 3. Intestinal epithelial cells express IFNkR. Human duodenal biopsies (see Fig. 1) and T84 cells were processed for immune staining. IFNkR was stained in green. The nuclei were stained in red. Sample annotation: (A) food allergy; (B) non-food allergy; (C) isotype control; (D) T84 cells; (E) isotype control. The data represent three experiments. Original magnification: 630. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
T84 cells and the intestinal epithelial layer. The results indicate that the Eo-derived IFNk induces the intestinal epithelial barrier dysfunction via causing the intestinal epithelial cell apoptosis.
3.5. Eo-derived IFNk plays an important role in the initiation of food allergen-related Th2 polarization in the intestine Finally we examined the role of Eo-derived IFNk in the pathogenesis of aberrant Th2 polarization. With the same mouse model in Fig. 3, we induced eosinophilia in the intestine. The mice were gavage-fed with a food allergen together with i.p. injection with LPS or PGN. As shown in Fig. 5, the high levels of IL-4 and OVA-specific IgE were detected in the intestinal protein extracts; high frequencies of Eos and mast cells were observed in the intestinal sections; and high frequencies of the OVA-specific Th2 cells were also detected in the intestine. On the other hand, the serum levels of IFNk were increased upon the treatment with LPS or PGN; the frequencies of Th1 cells and Tregs in the intestine were suppressed. Mice pretreated with anti-IFNk antibody, were not induced the aberrant Th2 polarization in the intestine; pretreatment with isotype IgG antibody did not block the aberrant Th2 polarization in the intestine. Mice with eosinophil depleted did not show any signs of Th2 inflammation in the intestine.
4. Discussion The present paper reports a novel finding that the intestinal Eos express IFNk. Microbial products, LPS and PGN, can activate Eos to release the IFNk. Intestinal epithelial cells express the IFNkR; the Eo-derived IFNk can compromise the intestinal epithelial barrier function by inducing the intestinal epithelial cell apoptosis. Concurrent exposure to microbial products and food antigens can induce the aberrant Th2 polarization in the intestine of mice that can be blocked by administration with anti-IFNk antibody.
The integrated intestinal epithelial barrier plays an important role in maintaining the homeostasis in the intestine. A number of intestinal disorders are related with the epithelial barrier dysfunction, such as food allergy and inflammatory bowel disease [15]. Our previous study revealed that the Th2 cytokine IL-4 could compromise intestinal epithelial barrier function [16]. Others found that IL-13 also induced intestinal epithelial barrier dysfunction [17]. In addition, other cytokines, such as IFNc [18], tumor necrosis factor alpha [19], and many others [20,21], could also affect the epithelial barrier function. The present study has revealed a previous undescribed factor, the IFNk, which also can cause intestinal epithelial barrier dysfunction. This cytokine exists in normal Eos, even under non-allergy conditions. Most of the previously reported factors, such as tumor necrosis factor alpha, involved in damaging the epithelial barrier function are also inflammatory mediators [22]; whether these factors are the results from epithelial barrier dysfunction-induced inflammation, or the pre-existed inflammation produces the cytokines, the latter in turn induces the epithelial barrier dysfunction, is not fully understood. In the present study, we found that mice with eosinophilia, which was caused by parasite infections, still maintained the normal intestinal epithelial barrier function. The mice had epithelial barrier dysfunction only occurred on exposing microbial products, abundant IFNk was released from Eos that caused the epithelial barrier permeability to increase. Of course, there are a number of other chemical mediators in Eos apart from IFNk; whether those mediators are also involved in the epithelial barrier dysfunction observed in the present study? The present data show that the increase in epithelial barrier permeability can be blocked by anti-IFNk antibody, demonstrating that IFNk is responsible for the epithelial barrier dysfunction in this experimental system. We also used microbial products, LPS and PGN, in the present study; these TLR ligands may not damage the epithelial barrier function directly because the intestinal epithelia are tolerant to these commensal bacterial products under the physiological condition [23,24]. Furthermore, upon exposure to
190
S.-H. He et al. / Cytokine 58 (2012) 186–192
Fig. 4. Eo-derived IFNk compromises the intestinal epithelial barrier integrity. T84 monolayers were cultured in transwells with Eos in the basal chambers in the presence of LPS or PGN. (A) Bars indicate the transepithelial electric resistance (TER) of T84 monolayers in response to LPS or PGN stimulation. (B) Bars indicate levels of horseradish peroxidase (HRP; using as a permeability tracer) in the culture medium in the basal chambers of transwells. (C and D) Bars indicate the levels of OVA (using as a tracer for assessing the epithelial permeability in vivo) in the intestinal tissue of mice treated with OVA/LPS (C) or OVA/PGN (D). (E) Bars indicate the frequency of the apoptotic T84 cells in (A) and (B), which was determined by flow cytometry. (F) The representative confocal images show apoptotic epithelial cells (F1, in green) and Eos (in blue) in the intestine of mice in (C) and (D). F2 is an isotype control. Original magnification: 200. Each experimental group consisted of six mice. The T84 cell experiments were performed three times. ⁄Pretreatment with anti-IFNk antibody (500 ng/ml). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
PGN, a ubiquitin editing E3 enzyme, the A20, is upregulated in the cells; the A20 in turn promotes tolerance to LPS in enterocytes [25,26]. The present data indicate that Eo-derived IFNk can compromise the epithelial barrier’s integrity via inducing epithelial cell apoptosis. The results are in line with previous observations in which we found that IFNk could cause CD3+ cell apoptosis [6]; others indicated that IFNk could induce cancer cell death via apoptotic pathway [7]. The epithelial cell apoptosis results in epithelial barrier defects and increase in epithelial barrier permeability to macromolecular antigens, such as OVA or HRP (their molecular weight is over 40 kDa), as shown by the sharply down-regulated TER of T84 monolayers and increase in HRP flux in the present data. Thus, the absorbed specific antigen, OVA, as shown by the in vivo data, is assumed being absorbed via the defect of epithelial layer caused by apoptosis. Under the physiological conditions, the ingested proteins are digested by enzymes in the intestine before they are absorbed into the intestinal tissue. The integral epithelial barrier restricts the absorption of macromolecular protein and large peptides; the absorbed nutrients usually have the least antigenicity. Upon exposure to Eo-derived IFNk, a large quantity of macromolecular protein passed the epithelial barrier and got into the intestinal tissue as shown by
the present data. The antigens then have the opportunity to contact the immune cells to initiate immune responses. The concurrently presented-microbial products, LPS or PGN, can speed up the immune response, or skew the immune response to aberrant Th2 polarization, as we recently observed that PGN could increase the T cell immunoglobulin mucin domain 4 (TIM-4) in dendritic cells and macrophages (data not shown); TIM-4 can induce Th2 polarization [27,28]. Indeed, employing a mouse model with eosinophilia, after exposed the mice to microbial products LPS or PGN, we did observe the skewed antigen-specific Th2 polarization and Th2 pattern inflammation in the intestine. Based on the present data and literature review, we may envisage a scenario like this: Under the condition of eosinophilia, such as the parasite infections, the eosinophils deposit a large quantity of IFNk in the body, as shown by the present data that normal Eos also have detectable IFNk. When the Eos encounter microbial stimuli, such as under psychological stress as we reported before [1,2], microbial products, such as LPS or PGN, cross the damaged epithelial barrier to enter deep tissues where the Eos disperse everywhere. The microbial products activate the TLRs on Eos to cause Eos to release IFNk, the latter causes epithelial barrier dysfunction resulting in proteins or large peptides with good antigenicity to be absorbed. The absorbed antigens and microbial
S.-H. He et al. / Cytokine 58 (2012) 186–192
191
Fig. 5. Eo-derived IFNk facilitates the aberrant Th2 polarization in the intestine. Mice with eosinophilia and OVA-sensitization were gavage-fed with OVA together with LPS or PGN i.p. as described in Supplementary materials. (A and B) Bars indicate the levels of and IL-4 (A) OVA specific IgE (sIgE) (B) in the protein extracts of the jejunum. (C) Bars indicate the rate of jejunal T cell proliferation that was assessed by the CFSE dilution. (D) Bars indicate the frequency of IL-4+ T cells among the proliferated cells in (C). (E) The bars indicate the serum levels of IFNk. (F) The LPMC were prepared and analyzed by flow cytometry. The bars indicate the frequencies of Treg or Th1 cells in the LPMC of mice treated with LPS or PGN. Each group consisted of six mice. M, mice were treated with microbial product (LPS or PGN, 1 lg/mouse, i.p.). a (or ia), mice were pretreated with anti-IFNk antibody (500 ng/mouse, i.p.) (or isotype IgG) 30 min prior to each challenge with antigen. Eo-, mice with eosinophil depleted. The data were presented as mean ± SD. ⁄p < 0.05, compared with the saline group. #p < 0.05, compared with the OVA-group.
products collaborate together to initiate the aberrant Th2 polarization and to further develop into allergic disorders such as food allergy. Author contributions S.-H.H., Z.L., C.-H.S., H.Z., W.M., L.-F.Z. and T.M. collected samples, performed experiments and revised the manuscript. S.-H.H. and P.-C.Y. designed the project, analyzed experimental data and wrote the manuscript. Acknowledgments This study was supported by grants from the Canadian Institutes of Health Research (Nos. 191063, 220058 and 177843), Natural Sciences, Engineering Research Council of Canada (NSERC, No. 371268), the Major State Basic Research Program of China (973 Program) (No. 2007CB512400) and the National Natural Science Foundation of China (Nos. 30972714 and 81030054). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.cyto.2012.01.003.
References [1] Zareie M, Johnson-Henry K, Jury J, Yang PC, Ngan BY, McKay DM, et al. Probiotics prevent bacterial translocation and improve intestinal barrier function in rats following chronic psychological stress. Gut 2006;55:1553–60. [2] Yang PC, Jury J, Soderholm JD, Sherman PM, McKay DM, Perdue MH. Chronic psychological stress in rats induces intestinal sensitization to luminal antigens. Am J Pathol 2006;168:104–14. [3] Collier-Hyams L, Neish A. Intestinal epithelial barrier and mucosal immunity. Cell Mol Life Sci 2005;62:1339–48. [4] Maresca M, Yahi N, YounFs-Sakr L, Boyron M, Caporiccio B, Fantini J. Both direct and indirect effects account for the pro-inflammatory activity of enteropathogenic mycotoxins on the human intestinal epithelium: stimulation of interleukin-8 secretion, potentiation of interleukin-1[beta] effect and increase in the transepithelial passage of commensal bacteria. Toxicol Appl Pharmacol 2008;228:84–92. [5] Vickery BP, Chin S, Burks AW. Pathophysiology of food allergy. Pediatr Clin North Am 2011;58:363–76. [6] He SH, Chen X, Song CH, Liu ZQ, Zhou LF, Ma WJ, et al. Interferon-[lambda] mediates oral tolerance and inhibits antigen-specific, T-helper 2 cell-mediated inflammation in mouse intestine. Gastroenterology 2011;141:249–58. [7] Li Q, Kawamura K, Ma G, Iwata F, Numasaki M, Suzuki N, et al. Interferon[lambda] induces G1 phase arrest or apoptosis in oesophageal carcinoma cells and produces anti-tumour effects in combination with anti-cancer agents. Eur J Cancer 2010;46:180–90. [8] Srinivas S, Dai J, Eskdale J, Gallagher GE, Megjugorac NJ, Gallagher G. Interferonlambda1 (interleukin-29) preferentially down-regulates interleukin-13 over other T helper type 2 cytokine responses in vitro. Immunology 2008;125:492–502. [9] Mennechet FJD, Uze G. Interferon-lambda-treated dendritic cells specifically induce proliferation of FOXP3-expressing suppressor T cells. Blood 2006;107: 4417–23.
192
S.-H. He et al. / Cytokine 58 (2012) 186–192
[10] Pott J, Mahlakoiv T, Mordstein M, Duerr CU, Michiels T, Stockinger S, et al. IFNlambda determines the intestinal epithelial antiviral host defense. Proc Natl Acad Sci USA 2011;108:7944–9. [11] Wang J, Oberley-Deegan R, Wang S, Nikrad M, Funk CJ, Hartshorn KL, et al. Differentiated human alveolar type II cells secrete antiviral IL-29 (IFNlambda1) in response to influenza A infection. J Immunol 2009;182:1296–304. [12] Remedios M, Jones D, Kerlin P. Eosinophilic oesophagitis: epidemiology, pathogenesis and management. Drugs 2011;71:527–40. [13] Plotz SG, Lentschat A, Behrendt H, Plotz W, Hamann L, Ring J, et al. The interaction of human peripheral blood eosinophils with bacterial lipopolysaccharide is CD14 dependent. Blood 2001;97:235–41. [14] Li Y, Efferson C, Ramesh R, Peoples G, Hwu P, Ioannides C. A peptidoglycan monomer with the glutamine to serine change and basic peptides bind in silico to TLR-2 (403–455). Cancer Immunol Immunother 2011;60:515–24. [15] Lamousé-Smith ES, Furuta GT. Eosinophils in the gastrointestinal tract. Curr Gastroenterol Rep 2006;8:390–5. [16] Di Leo V, Yang PC, Berin MC, Perdue MH. Factors regulating the effect of IL-4 on intestinal epithelial barrier function. Int Arch Allergy Immunol 2002;129: 219–27. [17] Heller F, Florian P, Bojarski C, Richter J, Christ M, Hillenbrand B, et al. Interleukin-13 is the key effector Th2 cytokine in ulcerative colitis that affects epithelial tight junctions, apoptosis, and cell restitution. Gastroenterology 2005;129:550–64. [18] Smyth D, Phan V, Wang A, McKay DM. Interferon-[gamma]-induced increases in intestinal epithelial macromolecular permeability requires the Src kinase Fyn. Lab Invest 2011;91:764–77. [19] Mashukova A, Wald FA, Salas PJ. Tumor necrosis factor alpha and inflammation disrupt the polarity complex in intestinal epithelial cells by a posttranslational mechanism. Mol Cell Biol 2011;31:756–65.
[20] Groschwitz KR, Hogan SP. Intestinal barrier function: molecular regulation and disease pathogenesis. J Allergy Clin Immunol 2009;124: 3–20. [21] Keita A, Soderholm JD. The intestinal barrier and its regulation by neuroimmune factors. Neurogastroenterol Motil 2010;22:718–33. [22] Ye D, Ma I, Ma TY. Molecular mechanism of tumor necrosis factor-alpha modulation of intestinal epithelial tight junction barrier. Am J Physiol Gastrointest Liver Physiol 2006;290:G496–504. [23] Lotz M, Gütle D, Walther S, Ménard S, Bogdan C, Hornef MW. Postnatal acquisition of endotoxin tolerance in intestinal epithelial cells. J Exp Med 2006;203:973–84. [24] Melmed G, Thomas LS, Lee N, Tesfay SY, Lukasek K, Michelsen KS, et al. Human intestinal epithelial cells are broadly unresponsive to toll-like receptor 2dependent bacterial ligands: implications for host–microbial interactions in the gut. J Immunol 2003;170:1406–15. [25] Wang J, Ouyang Y, Guner Y, Ford HR, Grishin AV. Ubiquitin-editing enzyme A20 promotes tolerance to lipopolysaccharide in enterocytes. J Immunol 2009;183:1384–92. [26] Hitotsumatsu O, Ahmad RC, Tavares R, Wang M, Philpott D, Turer EE, et al. The ubiquitin-editing enzyme A20 restricts nucleotide-binding oligomerization domain containing 2-triggered signals. Immunity 2008;28:381–90. [27] Yang PC, Xing Z, Berin CM, Soderholm JD, Feng BS, Wu L, et al. TIM-4 expressed by mucosal dendritic cells plays a critical role in food antigenspecific Th2 differentiation and intestinal allergy. Gastroenterology 2007;133:1522–33. [28] Meyers JH, Chakravarti S, Schlesinger D, Illes Z, Waldner H, Umetsu SE, et al. TIM-4 is the ligand for TIM-1, and the TIM-1-TIM-4 interaction regulates T cell proliferation. Nat Immunol 2005;6:455–64.
Contents lists available at SciVerse ScienceDirect
Cytokine journal homepage: www.elsevier.com/locate/issn/10434666
Eosinophil-derived interferon-lambda contributes to initiation of allergen-related inflammation in the intestine Shao-Heng He a,⇑,1, Chun-Hua Song b,1, Zhigang Liu c, Huiyun Zhang a, Wenjing Ma a, Lin-Fu Zhou a, Tahrin Mahmood b, Ping-Chang Yang b,⇑ a b c
Clinical Experimental Center, The First Affiliated Hospital, Nanjing Medical University, Nanjing 210029, China Department of Pathology & Molecular Medicine, McMaster University, Hamilton, ON, Canada Allergy and Immunology Institute, School of Medicine, Shenzhen University, Shenzhen, China
a r t i c l e
i n f o
Article history: Received 9 November 2011 Received in revised form 22 December 2011 Accepted 9 January 2012 Available online 4 February 2012 Keywords: T helper 2 response Eosinophil Interferon lambda Epithelial barrier Food allergy
a b s t r a c t Background and aims: Epithelial barrier dysfunction plays a critical role in the initiation of a number of immune diseases; the causative factors are not fully understood. The present study aimed to elucidate the mechanism by which the eosinophil-derived interferon (IFN)-lambda induced the gut epithelial barrier dysfunction. Methods: The duodenal biopsies were obtained from patients with or without food allergies. The eosinophils and IFNk expression were observed by immune staining. Intestinal epithelial cell line, T84 cells, and a mouse model were employed to observe the effect of IFNk on the epithelial barrier function and the initiation of skewed T helper (Th)2 polarization in the mouse intestine. Results: IFNk expression was observed in over 80% human eosinophils of the subjects with or without food allergies. Exposure to microbial products, lipopolysaccharide or peptidoglycan, could induce eosinophils to release IFNk. Exposure to IFNk could induce intestinal epithelial barrier dysfunction via inducing the epithelial cell apoptosis. Concurrent exposure to microbial products and food antigens could induce aberrant antigen specific Th2 polarization and Th2 pattern inflammation in the intestine. Conclusions: Eosinophils express IFNk that can induce intestinal epithelial barrier dysfunction and promotes the initiation of the aberrant Th2 polarization in the intestine. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction The antigen absorption in the intestinal mucosa is an essential step to initiate immune responses in the intestine. Under physiological conditions, the absorption of antigens is restricted by the integrated epithelial barrier on the intestinal mucosal surface. However, the integration of the intestinal epithelial barrier can be compromised by a number of factors, such as under psychological stress [1,2], or severe microbial infection [3,4], etc. These conditions may cause the tight junction open, increase in the intracellular permeability, or epithelial cell apoptosis resulting in the physical defect in the epithelial barrier; yet, the mechanisms underlying these conditions are not fully understood. Food allergy indicates a condition that the immune system in the intestine aberrantly responds to normal food components.
⇑ Corresponding authors. Address: Room T3303, 50 Charlton Ave. East, Hamilton, ON, Canada L8N 4A6. Tel.: +1 905 522 1155x32934; fax: +1 905 540 6593 (P.-C. Yang), tel./fax: +86 25 83718836 (S.-H. He). E-mail addresses: [email protected] (S.-H. He), [email protected] (P.-C. Yang). 1 These authors contributed equally to this work. 1043-4666/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.cyto.2012.01.003
The pathological feature of food allergy is a skewed T helper (Th)2 polarization, elevated levels of antigen specific IgE, mastocytosis and eosinophilia in the intestinal mucosa [5]. The initiation of the skewed Th2 polarization is not fully understood yet; multiple factors may be involved in the induction of food allergy. Interferon (IFN)k is a newly described cytokine, including IFNk1 [interleukin (IL)-29], IFNk2 (IL-28A) and IFNk3 (IL-28B) three subtypes; the IFNk1 is only expressed in human. IFNk is actively involved in the immune regulation, such as inducing target cell apoptosis [6,7], suppressing IL-13 expression [8], facilitating regulatory T cell development [6,9] and fighting against viral infections. It has been found that IFNk can be produced by several cells types, including epithelial cells [10], dendritic cells [6] and macrophages [11]. It is still not fully understood how the production of IFN initiated by those cells. Eosinophils (Eos) are an inflammatory cell type. The eosinophilia is found in a number of diseases, such as allergic disorders, parasite infections and some special inflammations like eosinophilic esophagitis [12]. Apart from producing a number of chemical mediators, such as the major basic protein (MBP), eosinophil cationic protein, eosinophil peroxidase, etc., we also noted that Eos also produced IFNk in our previous study ([6] data not shown);
S.-H. He et al. / Cytokine 58 (2012) 186–192
187
Fig. 1. Eos express IFNk in the intestine. Duodenal biopsies were processed for immunohistochemistry. The representative confocal images show that the Eo marker, MBP, was stained in green; the IFNk was stained in blue. The light blue color (pointed by arrows) was merged by the colors of green and blue. The nuclei were stained in red. (A) Samples were obtained from 15 patients with food allergy. (B) Samples were obtained from 15 non-food allergy patients. (C) An isotype control. (D) The bars indicate the double positive cell counts (mean ± SD). ⁄p < 0.01, compared with food allergy (FA) group. nFA, non-FA. Original magnification: 200. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
the phenomenon has not been reported yet as we checked the literature throughout the PubMed database. Since Eos are one of the effector cell types in allergic diseases, Eos express Toll like receptor (TLR)2 and TLR4 [13], we postulated that the Eo-derived IFNk might be involved in the initiation of the skewed Th2 polarization in the intestine. The present data show that upon exposure to the ligands of TLR2 or TLR4, Eos release IFNk; the latter affects the integrity of intestinal epithelial barrier function that contributes to the initiation of aberrant Th2 polarization and Th2 pattern inflammation in the intestine. 2. Materials and methods 2.1. Patients Thirty patients with peptic ulcer with or without food allergies were included in the present study. The diagnosis of food allergy is described in Supplementary materials. The patients did not use any anti-allergy drugs at least 1 month before the study. The demographic data of patients are presented in Supplementary Table 1.
2.4. Immunohistochemistry and TUNEL staining Duodenal biopsies were frozen in liquid nitrogen and stored at 80 °C until use. Cryosections were prepared and fixed with cold acetone for 20 min. Sections were blocked by 1% bovine serum albumin (BSA) for 30 min and incubated with goat anti-human IFNk antibody (1:200) and mouse anti-MBP antibody (1:300), or isotype IgG (using as negative control) overnight at 4 °C. After washing, sections were incubated with Cy5-labeled anti-goat and FITC-labeled anti-mouse second antibodies (1:300) for 1 h at room temperature. After washing, sections were mounted with cover slips and observed under a confocal microscope. The positive cells were counted in 20 fields (200) per samples. For T84 cell staining, the cell suspension was smeared onto glass slides, dried overnight. The rest procedures were the same as above. To avoid the observer bias, the sections were coded; the observer was not aware of the codes. After the MBP labeling, some sections were incubated for 1 h in freshly prepared TUNEL mixture containing TUNEL plus fluorescein tag in a dark humidified chamber at 37 °C. After washing, sections were mounted with coverslips and viewed under a confocal microscope.
2.2. Duodenal biopsy 2.5. Detecting intestinal epithelial barrier function in vivo Duodenal biopsies were taken under endoscopy for a purpose of ruling out malignant disorders in the duodenum. One piece of biopsies was taken from the edge of the ulcer; another piece was taken from non-ulcer mucosa. The procedures using human tissue in the study were approved by the Research Ethic Committee at Nanjing Medical University. Informed written consent was obtained from each subject.
BALB/c mice with eosinophilia were gavage-fed with ovalbumin (OVA) 5 mg/mouse in 0.3 ml saline. Mice were also i.p. injected with lipopolysaccharide (LPS; see Fig. 3C for dosage) or peptidoglycan (PGN; see Fig. 3D for dosage). Mice were sacrificed 6 h later. The jejunum segments were excised to extract the total proteins. The proteins were analyzed by Western blotting for the contents of the OVA.
2.3. Enzyme-linked immunoassay (ELISA) 2.6. Eosinophil depletion Levels of serum specific IgE, intestinal IL-4 and IFNk in tissue extracts and culture supernatant were measured by ELISA with commercial reagent kits following the manufacturer’s instruction.
Rats were i.p. injection with anti-CCR3 antibody (0.5 mg/rat) on day 3, 8 and 11, respectively. As shown by peripheral blood smear
188
S.-H. He et al. / Cytokine 58 (2012) 186–192
85% ± 11.8% of the total MBP+ cells in food allergy group and 82.1% ± 14.8% of the total MBP+ cells in non-food allergy group (p > 0.05; Fig. 1). In addition, similar results were also observed in human tissues of the skin, tonsil, colon, lung, kidney and stomach (data not shown). To confirm the data that Eos express IFNk, we infected mice with parasites to induce the eosinophilia in the intestine. The nucleated cells were isolated from the jejunum and examined by flow cytometry. The results showed that the rate of MBP+ IFNk+ cells was 92.5% ± 15.6% in the total MBP+ cells from six mice with parasite infection and 91.1% ± 16.6% in naïve mice. The data indicate that the number of Eo is significantly more in the duodenal mucosa of patients with food allergy than those of non-food allergy. Over 80% Eos in human and over 90% Eos in mice express IFNk. 3.2. Eos release IFNk upon activation by TLRs Previous reports indicate that Eos express TLR2 and TLR4 [13]. We next examined the expression of TLRs on the Eo cell line, EoL1 cells. As shown by flow cytometry, over 90% EoL1 cells express both TLR2 and TLR4 (Fig. 2A and B). The EoL1 cells also expressed IFNk (Fig. 2C and D). We then added the ligands of TLR2 (peptidoglycan; PGN; 100 ng/ml) [14] and TLR4 (lipopolysaccharide; LPS; 100 ng/ml) to activate the EoL1 cells in culture for 3 days. As assessed by ELISA, both TLR2 and TLR4 ligands could respectively induce the release of IFNk from EoL1 cells (Fig. 2E). 3.3. Intestinal epithelial cells express the IFNk receptors (IFNkR)
Fig. 2. EoL1 cells release IFNk upon TLR ligand stimulation. (A and B) The dot plots show TLR2 and TLR4 expressing EoL1 cells. (B) An isotype control. (C and D) The histograms indicate the expression of IFNk by the EoL1 cells in (A) and (B) (pointed by arrows). (E) The bars indicate the levels of IFNk in culture supernatants of EoL1 cells stimulated by LPS of PGN. The data were expressed as mean ± SD from three experiments. ⁄p < 0.01, compared with the dose ‘‘0’’ group.
and jejunal sections, eosinophils were observed in naïve control rats; no eosinophils were observed in rats treated with anti-CCR3 antibody. 2.7. Statistics All values were expressed as the means ± SD of at least three independent experiments. The values were analyzed using the two-tailed unpaired Student’s t-test when the data consisted of two groups or by ANOVA when three or more groups were compared. p < 0.05 was accepted as statistically significant. Other experimental procedures were listed in Supplementary materials. 3. Results 3.1. Eos express IFN in the intestine In the first attempt, we observed the expression of IFNk by Eos in the duodenal biopsies. As shown by immunohistochemistry, Eos were observed in the duodenal biopsy sections of the patients with or without food allergies. The number of Eos was significantly more in the patients with food allergies than that in non-food allergy (p < 0.01). Three types of positive cells, MBP+, IFNk+ and MBP+ IFNk+, were observed. The rate of MBP+ IFNk+ cells was
Eos are dispersed immediately under the subepithelial region as well as in the epithelial layer. Such an anatomical relation between Eos and the epithelium implies that the Eo-derived IFNk may affect the epithelial cells. By immunohistochemistry, the expression of IFNkR was observed on the epithelial cells of human duodenal biopsies as well as on the human intestinal cell line, T84 cells (Fig. 3). 3.4. Eo-derived IFNk induces gut epithelial barrier dysfunction The data in Figs. 2 and 3 imply that Eo-derived IFNk may affect the intestinal epithelial barrier function. To test the hypothesis, we cultured the human intestinal epithelial cell line, T84 cell monolayers, in transwells to confluence. EoL1 cells were added to the basal chambers of transwells in the presence of LPS or PGN. The results showed that upon the exposure to LPS or PGN, the transepithelial electric resistance (TER) was markedly dropped and the permeability to HRP was significantly increased, which could be blocked by addition of the anti-IFNk antibody (Fig. 4A and B). The data imply that the Eo-derived IFNk can compromise the intestinal epithelial barrier function. To further confirm the results in Fig. 4A, we induced the eosinophilia in mouse intestine by infecting mice with parasites. The mice with eosinophilia were injected with LPS or PGN via i.p. to induce the Eos to release IFNk; the mice were concurrently gavage-fed with OVA to determine the epithelial barrier permeability. As shown by Western blotting (Fig. 4C and D), the contents of OVA in the protein extracts of intestinal tissue were significantly increased in a TLR ligand dose-dependent manner, which could be blocked by pretreatment with the anti-IFNk antibody, but not by pretreatment with an isotype antibody (data not shown). The results indicate that Eo-derived IFNk can cause intestinal epithelial barrier dysfunction. To find the mechanism by which Eo-derived IFNk induces intestinal epithelial barrier dysfunction, we examined the T84 cells in Fig. 3 and the intestinal epithelial cells of mice in Fig. 4. The results showed that abundant apoptotic cells were detected among the
S.-H. He et al. / Cytokine 58 (2012) 186–192
189
Fig. 3. Intestinal epithelial cells express IFNkR. Human duodenal biopsies (see Fig. 1) and T84 cells were processed for immune staining. IFNkR was stained in green. The nuclei were stained in red. Sample annotation: (A) food allergy; (B) non-food allergy; (C) isotype control; (D) T84 cells; (E) isotype control. The data represent three experiments. Original magnification: 630. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
T84 cells and the intestinal epithelial layer. The results indicate that the Eo-derived IFNk induces the intestinal epithelial barrier dysfunction via causing the intestinal epithelial cell apoptosis.
3.5. Eo-derived IFNk plays an important role in the initiation of food allergen-related Th2 polarization in the intestine Finally we examined the role of Eo-derived IFNk in the pathogenesis of aberrant Th2 polarization. With the same mouse model in Fig. 3, we induced eosinophilia in the intestine. The mice were gavage-fed with a food allergen together with i.p. injection with LPS or PGN. As shown in Fig. 5, the high levels of IL-4 and OVA-specific IgE were detected in the intestinal protein extracts; high frequencies of Eos and mast cells were observed in the intestinal sections; and high frequencies of the OVA-specific Th2 cells were also detected in the intestine. On the other hand, the serum levels of IFNk were increased upon the treatment with LPS or PGN; the frequencies of Th1 cells and Tregs in the intestine were suppressed. Mice pretreated with anti-IFNk antibody, were not induced the aberrant Th2 polarization in the intestine; pretreatment with isotype IgG antibody did not block the aberrant Th2 polarization in the intestine. Mice with eosinophil depleted did not show any signs of Th2 inflammation in the intestine.
4. Discussion The present paper reports a novel finding that the intestinal Eos express IFNk. Microbial products, LPS and PGN, can activate Eos to release the IFNk. Intestinal epithelial cells express the IFNkR; the Eo-derived IFNk can compromise the intestinal epithelial barrier function by inducing the intestinal epithelial cell apoptosis. Concurrent exposure to microbial products and food antigens can induce the aberrant Th2 polarization in the intestine of mice that can be blocked by administration with anti-IFNk antibody.
The integrated intestinal epithelial barrier plays an important role in maintaining the homeostasis in the intestine. A number of intestinal disorders are related with the epithelial barrier dysfunction, such as food allergy and inflammatory bowel disease [15]. Our previous study revealed that the Th2 cytokine IL-4 could compromise intestinal epithelial barrier function [16]. Others found that IL-13 also induced intestinal epithelial barrier dysfunction [17]. In addition, other cytokines, such as IFNc [18], tumor necrosis factor alpha [19], and many others [20,21], could also affect the epithelial barrier function. The present study has revealed a previous undescribed factor, the IFNk, which also can cause intestinal epithelial barrier dysfunction. This cytokine exists in normal Eos, even under non-allergy conditions. Most of the previously reported factors, such as tumor necrosis factor alpha, involved in damaging the epithelial barrier function are also inflammatory mediators [22]; whether these factors are the results from epithelial barrier dysfunction-induced inflammation, or the pre-existed inflammation produces the cytokines, the latter in turn induces the epithelial barrier dysfunction, is not fully understood. In the present study, we found that mice with eosinophilia, which was caused by parasite infections, still maintained the normal intestinal epithelial barrier function. The mice had epithelial barrier dysfunction only occurred on exposing microbial products, abundant IFNk was released from Eos that caused the epithelial barrier permeability to increase. Of course, there are a number of other chemical mediators in Eos apart from IFNk; whether those mediators are also involved in the epithelial barrier dysfunction observed in the present study? The present data show that the increase in epithelial barrier permeability can be blocked by anti-IFNk antibody, demonstrating that IFNk is responsible for the epithelial barrier dysfunction in this experimental system. We also used microbial products, LPS and PGN, in the present study; these TLR ligands may not damage the epithelial barrier function directly because the intestinal epithelia are tolerant to these commensal bacterial products under the physiological condition [23,24]. Furthermore, upon exposure to
190
S.-H. He et al. / Cytokine 58 (2012) 186–192
Fig. 4. Eo-derived IFNk compromises the intestinal epithelial barrier integrity. T84 monolayers were cultured in transwells with Eos in the basal chambers in the presence of LPS or PGN. (A) Bars indicate the transepithelial electric resistance (TER) of T84 monolayers in response to LPS or PGN stimulation. (B) Bars indicate levels of horseradish peroxidase (HRP; using as a permeability tracer) in the culture medium in the basal chambers of transwells. (C and D) Bars indicate the levels of OVA (using as a tracer for assessing the epithelial permeability in vivo) in the intestinal tissue of mice treated with OVA/LPS (C) or OVA/PGN (D). (E) Bars indicate the frequency of the apoptotic T84 cells in (A) and (B), which was determined by flow cytometry. (F) The representative confocal images show apoptotic epithelial cells (F1, in green) and Eos (in blue) in the intestine of mice in (C) and (D). F2 is an isotype control. Original magnification: 200. Each experimental group consisted of six mice. The T84 cell experiments were performed three times. ⁄Pretreatment with anti-IFNk antibody (500 ng/ml). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
PGN, a ubiquitin editing E3 enzyme, the A20, is upregulated in the cells; the A20 in turn promotes tolerance to LPS in enterocytes [25,26]. The present data indicate that Eo-derived IFNk can compromise the epithelial barrier’s integrity via inducing epithelial cell apoptosis. The results are in line with previous observations in which we found that IFNk could cause CD3+ cell apoptosis [6]; others indicated that IFNk could induce cancer cell death via apoptotic pathway [7]. The epithelial cell apoptosis results in epithelial barrier defects and increase in epithelial barrier permeability to macromolecular antigens, such as OVA or HRP (their molecular weight is over 40 kDa), as shown by the sharply down-regulated TER of T84 monolayers and increase in HRP flux in the present data. Thus, the absorbed specific antigen, OVA, as shown by the in vivo data, is assumed being absorbed via the defect of epithelial layer caused by apoptosis. Under the physiological conditions, the ingested proteins are digested by enzymes in the intestine before they are absorbed into the intestinal tissue. The integral epithelial barrier restricts the absorption of macromolecular protein and large peptides; the absorbed nutrients usually have the least antigenicity. Upon exposure to Eo-derived IFNk, a large quantity of macromolecular protein passed the epithelial barrier and got into the intestinal tissue as shown by
the present data. The antigens then have the opportunity to contact the immune cells to initiate immune responses. The concurrently presented-microbial products, LPS or PGN, can speed up the immune response, or skew the immune response to aberrant Th2 polarization, as we recently observed that PGN could increase the T cell immunoglobulin mucin domain 4 (TIM-4) in dendritic cells and macrophages (data not shown); TIM-4 can induce Th2 polarization [27,28]. Indeed, employing a mouse model with eosinophilia, after exposed the mice to microbial products LPS or PGN, we did observe the skewed antigen-specific Th2 polarization and Th2 pattern inflammation in the intestine. Based on the present data and literature review, we may envisage a scenario like this: Under the condition of eosinophilia, such as the parasite infections, the eosinophils deposit a large quantity of IFNk in the body, as shown by the present data that normal Eos also have detectable IFNk. When the Eos encounter microbial stimuli, such as under psychological stress as we reported before [1,2], microbial products, such as LPS or PGN, cross the damaged epithelial barrier to enter deep tissues where the Eos disperse everywhere. The microbial products activate the TLRs on Eos to cause Eos to release IFNk, the latter causes epithelial barrier dysfunction resulting in proteins or large peptides with good antigenicity to be absorbed. The absorbed antigens and microbial
S.-H. He et al. / Cytokine 58 (2012) 186–192
191
Fig. 5. Eo-derived IFNk facilitates the aberrant Th2 polarization in the intestine. Mice with eosinophilia and OVA-sensitization were gavage-fed with OVA together with LPS or PGN i.p. as described in Supplementary materials. (A and B) Bars indicate the levels of and IL-4 (A) OVA specific IgE (sIgE) (B) in the protein extracts of the jejunum. (C) Bars indicate the rate of jejunal T cell proliferation that was assessed by the CFSE dilution. (D) Bars indicate the frequency of IL-4+ T cells among the proliferated cells in (C). (E) The bars indicate the serum levels of IFNk. (F) The LPMC were prepared and analyzed by flow cytometry. The bars indicate the frequencies of Treg or Th1 cells in the LPMC of mice treated with LPS or PGN. Each group consisted of six mice. M, mice were treated with microbial product (LPS or PGN, 1 lg/mouse, i.p.). a (or ia), mice were pretreated with anti-IFNk antibody (500 ng/mouse, i.p.) (or isotype IgG) 30 min prior to each challenge with antigen. Eo-, mice with eosinophil depleted. The data were presented as mean ± SD. ⁄p < 0.05, compared with the saline group. #p < 0.05, compared with the OVA-group.
products collaborate together to initiate the aberrant Th2 polarization and to further develop into allergic disorders such as food allergy. Author contributions S.-H.H., Z.L., C.-H.S., H.Z., W.M., L.-F.Z. and T.M. collected samples, performed experiments and revised the manuscript. S.-H.H. and P.-C.Y. designed the project, analyzed experimental data and wrote the manuscript. Acknowledgments This study was supported by grants from the Canadian Institutes of Health Research (Nos. 191063, 220058 and 177843), Natural Sciences, Engineering Research Council of Canada (NSERC, No. 371268), the Major State Basic Research Program of China (973 Program) (No. 2007CB512400) and the National Natural Science Foundation of China (Nos. 30972714 and 81030054). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.cyto.2012.01.003.
References [1] Zareie M, Johnson-Henry K, Jury J, Yang PC, Ngan BY, McKay DM, et al. Probiotics prevent bacterial translocation and improve intestinal barrier function in rats following chronic psychological stress. Gut 2006;55:1553–60. [2] Yang PC, Jury J, Soderholm JD, Sherman PM, McKay DM, Perdue MH. Chronic psychological stress in rats induces intestinal sensitization to luminal antigens. Am J Pathol 2006;168:104–14. [3] Collier-Hyams L, Neish A. Intestinal epithelial barrier and mucosal immunity. Cell Mol Life Sci 2005;62:1339–48. [4] Maresca M, Yahi N, YounFs-Sakr L, Boyron M, Caporiccio B, Fantini J. Both direct and indirect effects account for the pro-inflammatory activity of enteropathogenic mycotoxins on the human intestinal epithelium: stimulation of interleukin-8 secretion, potentiation of interleukin-1[beta] effect and increase in the transepithelial passage of commensal bacteria. Toxicol Appl Pharmacol 2008;228:84–92. [5] Vickery BP, Chin S, Burks AW. Pathophysiology of food allergy. Pediatr Clin North Am 2011;58:363–76. [6] He SH, Chen X, Song CH, Liu ZQ, Zhou LF, Ma WJ, et al. Interferon-[lambda] mediates oral tolerance and inhibits antigen-specific, T-helper 2 cell-mediated inflammation in mouse intestine. Gastroenterology 2011;141:249–58. [7] Li Q, Kawamura K, Ma G, Iwata F, Numasaki M, Suzuki N, et al. Interferon[lambda] induces G1 phase arrest or apoptosis in oesophageal carcinoma cells and produces anti-tumour effects in combination with anti-cancer agents. Eur J Cancer 2010;46:180–90. [8] Srinivas S, Dai J, Eskdale J, Gallagher GE, Megjugorac NJ, Gallagher G. Interferonlambda1 (interleukin-29) preferentially down-regulates interleukin-13 over other T helper type 2 cytokine responses in vitro. Immunology 2008;125:492–502. [9] Mennechet FJD, Uze G. Interferon-lambda-treated dendritic cells specifically induce proliferation of FOXP3-expressing suppressor T cells. Blood 2006;107: 4417–23.
192
S.-H. He et al. / Cytokine 58 (2012) 186–192
[10] Pott J, Mahlakoiv T, Mordstein M, Duerr CU, Michiels T, Stockinger S, et al. IFNlambda determines the intestinal epithelial antiviral host defense. Proc Natl Acad Sci USA 2011;108:7944–9. [11] Wang J, Oberley-Deegan R, Wang S, Nikrad M, Funk CJ, Hartshorn KL, et al. Differentiated human alveolar type II cells secrete antiviral IL-29 (IFNlambda1) in response to influenza A infection. J Immunol 2009;182:1296–304. [12] Remedios M, Jones D, Kerlin P. Eosinophilic oesophagitis: epidemiology, pathogenesis and management. Drugs 2011;71:527–40. [13] Plotz SG, Lentschat A, Behrendt H, Plotz W, Hamann L, Ring J, et al. The interaction of human peripheral blood eosinophils with bacterial lipopolysaccharide is CD14 dependent. Blood 2001;97:235–41. [14] Li Y, Efferson C, Ramesh R, Peoples G, Hwu P, Ioannides C. A peptidoglycan monomer with the glutamine to serine change and basic peptides bind in silico to TLR-2 (403–455). Cancer Immunol Immunother 2011;60:515–24. [15] Lamousé-Smith ES, Furuta GT. Eosinophils in the gastrointestinal tract. Curr Gastroenterol Rep 2006;8:390–5. [16] Di Leo V, Yang PC, Berin MC, Perdue MH. Factors regulating the effect of IL-4 on intestinal epithelial barrier function. Int Arch Allergy Immunol 2002;129: 219–27. [17] Heller F, Florian P, Bojarski C, Richter J, Christ M, Hillenbrand B, et al. Interleukin-13 is the key effector Th2 cytokine in ulcerative colitis that affects epithelial tight junctions, apoptosis, and cell restitution. Gastroenterology 2005;129:550–64. [18] Smyth D, Phan V, Wang A, McKay DM. Interferon-[gamma]-induced increases in intestinal epithelial macromolecular permeability requires the Src kinase Fyn. Lab Invest 2011;91:764–77. [19] Mashukova A, Wald FA, Salas PJ. Tumor necrosis factor alpha and inflammation disrupt the polarity complex in intestinal epithelial cells by a posttranslational mechanism. Mol Cell Biol 2011;31:756–65.
[20] Groschwitz KR, Hogan SP. Intestinal barrier function: molecular regulation and disease pathogenesis. J Allergy Clin Immunol 2009;124: 3–20. [21] Keita A, Soderholm JD. The intestinal barrier and its regulation by neuroimmune factors. Neurogastroenterol Motil 2010;22:718–33. [22] Ye D, Ma I, Ma TY. Molecular mechanism of tumor necrosis factor-alpha modulation of intestinal epithelial tight junction barrier. Am J Physiol Gastrointest Liver Physiol 2006;290:G496–504. [23] Lotz M, Gütle D, Walther S, Ménard S, Bogdan C, Hornef MW. Postnatal acquisition of endotoxin tolerance in intestinal epithelial cells. J Exp Med 2006;203:973–84. [24] Melmed G, Thomas LS, Lee N, Tesfay SY, Lukasek K, Michelsen KS, et al. Human intestinal epithelial cells are broadly unresponsive to toll-like receptor 2dependent bacterial ligands: implications for host–microbial interactions in the gut. J Immunol 2003;170:1406–15. [25] Wang J, Ouyang Y, Guner Y, Ford HR, Grishin AV. Ubiquitin-editing enzyme A20 promotes tolerance to lipopolysaccharide in enterocytes. J Immunol 2009;183:1384–92. [26] Hitotsumatsu O, Ahmad RC, Tavares R, Wang M, Philpott D, Turer EE, et al. The ubiquitin-editing enzyme A20 restricts nucleotide-binding oligomerization domain containing 2-triggered signals. Immunity 2008;28:381–90. [27] Yang PC, Xing Z, Berin CM, Soderholm JD, Feng BS, Wu L, et al. TIM-4 expressed by mucosal dendritic cells plays a critical role in food antigenspecific Th2 differentiation and intestinal allergy. Gastroenterology 2007;133:1522–33. [28] Meyers JH, Chakravarti S, Schlesinger D, Illes Z, Waldner H, Umetsu SE, et al. TIM-4 is the ligand for TIM-1, and the TIM-1-TIM-4 interaction regulates T cell proliferation. Nat Immunol 2005;6:455–64.